04:04:45 <ekcs> #startmeeting congressteammeeting
04:04:46 <openstack> Meeting started Fri Jun 28 04:04:45 2019 UTC and is due to finish in 60 minutes.  The chair is ekcs. Information about MeetBot at http://wiki.debian.org/MeetBot.
04:04:47 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
04:04:49 <openstack> The meeting name has been set to 'congressteammeeting'
04:06:13 <ekcs> hello! topics go here as usual: https://etherpad.openstack.org/p/congress-meeting-topics
04:08:04 <akhil_jain> Hi, ekcs
04:08:22 <ekcs> hello akhil_jain ! how’s it been?
04:08:46 <ekcs> and hello imsurit ! if you’re here to join the meeting, welcome!
04:08:52 <akhil_jain> Everything good, what about you?
04:11:21 <ekcs> im alright too!
04:12:20 <ekcs> a little bit of a stressful week but reasonably productive = )
04:12:57 <ekcs> ok I guess we’ll dive in then.
04:13:07 <ekcs> #topic intelligent overprovisioning for high availability
04:14:14 <ekcs> I just want to give a quick update to this use case I’ve been working on.
04:15:07 <ekcs> it’s mentioned in our talks in the denver summit, but I’m working on formalizing it for the self-healing SIG
04:15:19 <ekcs> and also fleshing out all the reference deployment setup.
04:17:20 <ekcs> the concept is to use predictive analytics (maybe monasca-analytics) to warn of possible failures. and based on those failure severetiy, likelihood, and the load on the affected services, decide how much to overprovision to guard against service disruption/degradation if the failure occurs.
04:18:26 <ekcs> so something like monasca analytics would provide the predictive data, and something like congress would combine with other data to make provisioning decisions.
04:19:11 <ekcs> one difficulty right now is to find someone who is familiar with the predictive piece to drive that. but currently i’m working on the congress side.
04:20:05 <akhil_jain> it sounds interesting
04:20:11 <ekcs> the numerical aggregations and calculations in the postgres engine for congress has been very helpful. will update as I progress.
04:20:49 <akhil_jain> are you docuenting it in self healing sig ?
04:21:10 <ekcs> yea I think it makes sense to document for self healing SIG and also auto-scaling SIG.
04:22:09 <akhil_jain> is patch up?
04:22:09 <ekcs> but also if I can find a monasca-analytics person or someone who is familiar with predictive analytics it may make sense to submit a talk for shanghai.
04:22:23 <ekcs> no not yet. will add you to review when I submit a patch though.
04:22:38 <akhil_jain> ok thanks
04:23:03 <ekcs> is that something you’d be interested in collaborating on for a presentation?
04:23:43 <akhil_jain> donot know of any such person, but i can be one by studying,
04:23:51 <akhil_jain> always ready:D
04:24:47 <ekcs> haha ok well we can talk more about level of interest and work.
04:24:50 <akhil_jain> or maybe asking witek can help here
04:25:09 <ekcs> anyway that’s all the uupdate from me on that topic.
04:25:42 <akhil_jain> ok great, thanks for updating
04:26:23 <ekcs> ok moving on then.
04:26:32 <ekcs> #topic summit presentations
04:27:06 <ekcs> on the same line of thought, just want to see if there are any ideas or discussions around talks or forums for the summit
04:29:27 <ekcs> or it’s something we can discuss more over email if they come up.
04:30:22 <akhil_jain> I am not sure about the way we can add idea of prometheus-openstack monitoring, I have written down rough abstract, will share it with you over mail by tomorrow
04:30:50 <ekcs> ok great!
04:31:34 <ekcs> ok moving on then.
04:31:43 <ekcs> #topic open discussion
04:31:57 <ekcs> not sure if we have anything else to talk about right now.
04:32:09 <akhil_jain> yea, nothing from my side
04:32:48 <ekcs> ok nothing for me either. I guess we can end then!
04:33:51 <akhil_jain> ok, later then
04:33:57 <ekcs> happy friday and have a great weekend!
04:34:11 <akhil_jain> Thanks and same to you, Bye!
04:34:27 <ekcs> bye
04:34:30 <ekcs> #endmeeting